docs: DOC-899: Keyed transpose user guide #7266
Conversation
| title: Keyed transpose | ||
| --- | ||
|
|
||
| This guide shows you how to use `keyedTranspose` to transform data from a long format to a wide format by pivoting column values into new column names. This is useful when you need to reshape data for analysis, reporting, or visualization. |
There was a problem hiding this comment.
Link to keyedTranspose reference doc.
|
|
||
| This guide shows you how to use `keyedTranspose` to transform data from a long format to a wide format by pivoting column values into new column names. This is useful when you need to reshape data for analysis, reporting, or visualization. | ||
|
|
||
| ## When to use keyedTranspose |
There was a problem hiding this comment.
IMO we should refer to it as a keyed transpose in headings and whatnot.
|
|
||
| ## When to use keyedTranspose | ||
|
|
||
| Use `keyedTranspose` when you need to: |
| - Each unique `Level` value (INFO, WARN, ERROR) becomes a column. | ||
| - The `Count` aggregation counts occurrences for each Date-Level combination. | ||
|
|
||
| ## Working with multiple row keys |
There was a problem hiding this comment.
| ## Working with multiple row keys | |
| ## Multiple row keys |
|
|
||
| Each unique combination of `Date` and `Server` creates a separate row in the output. | ||
|
|
||
| ## Using multiple aggregations |
There was a problem hiding this comment.
| ## Using multiple aggregations | |
| ## Multiple aggregations |
| | Single aggregation, single column-by | Value from column-by column | `INFO`, `WARN` | | ||
| | Multiple aggregations | Aggregation name + value | `Count_INFO`, `Sum_WARN` | | ||
| | Multiple column-by columns | Values joined with underscore | `INFO_10`, `WARN_20` | |
There was a problem hiding this comment.
Without examples to illustrate these, I don't know what this means.
| | Starts with number | Prefixed with `column_` | `123` → `column_123` | | ||
| | Duplicate names | Suffix added | `INFO`, `INFO2` | | ||
|
|
||
| ### Best practice: Sanitize data before transposing |
There was a problem hiding this comment.
I'm not sure I'd say this is a best practice. In cases of larger tables, mass string replace operations could be expensive.
A better way to go about this would be to show an example of everything in the table above. Each example can go into its own subsection. Either that, or you can present every item in the table above as a subsection with an example.
|
|
||
| ## Common patterns | ||
|
|
||
| ### Pattern 1: Sales by region and product |
There was a problem hiding this comment.
This feels more like a simple example than a common pattern.
| ) | ||
| ``` | ||
|
|
||
| ### Pattern 2: Time-series metrics |
There was a problem hiding this comment.
IMO presenting these as examples rather than "common patterns" would be beneficial. I don't see how these are common patterns, but they're obviously useful examples for learning.
There was a problem hiding this comment.
Same comments on Python that I made on Groovy (if they apply).
http://deephaven.atlassian.net/jira/software/projects/DOC/boards/8?selectedIssue=DOC-899